When sausage rolls go offline: The hidden cost of outages

Last week, I was with AI-outage-prediction company Hortium at the Retail Technology show at Olympia.

Why was that?

WhatApp, Instagram and a string of retailers including McDonalds and ultimate champion dealer of the sausage roll, Greggs, have all suffered outages over the last year that have stopped users using and buyers buying.Â

Last year, UK supermarkets Tesco and Sainsbury’s faced point-of-sale system outages, disrupting transactions.

What is going on?

Industry consensus suggests the average cost of retail downtime is around $5,600 per minute. These figures encompass not only lost sales but also diminished customer trust, brand damage, and operational recovery expenses, underscoring the high stakes involved in swift and effective outage management.

On average, resolving a critical system outage can take between four to eight hours (480 mins x $5600 = $2.68m), with more severe or complex issues requiring longer to address fully.

And it looks like they are going to be more.

As retailers continue to integrate more sophisticated technologies to enhance customer experiences, the complexity of managing systems increases.

Global technology services are ugly beasts that are getting harder to tame. 

The more companies rely on outsourced, cloud-based services while trying to manage complex legacy systems, the less visibility and control they have.

That also means diagnosing faults is taking longer – and trying to hold vendors accountable is extremely difficult. If you can prove responsibility, some vendors offer credits or financial compensation for downtime – but it’s not easy.

For global companies with complex architectures, AI plays a vital role in outage prevention.

It performs better than humans to predict failures and alert people.

It’s not a nice to have – if companies want less disruption, they need to use AI to prevent downtime.